MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <18895.18217.393755.824092@cargo.ozlabs.ibm.com>
Date: Sun, 29 Mar 2009 21:02:17 +1100
From: Paul Mackerras <paulus@samba.org>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@elte.hu>, linux-kernel@vger.kernel.org,
       Mike Galbraith <efault@gmx.de>, Arjan van de Ven <arjan@infradead.org>,
       Wu Fengguang <fengguang.wu@intel.com>,
       Eric Dumazet <dada1@cosmosbay.com>
Subject: Re: [PATCH 1/9] perf_counter: unify and fix delayed counter wakeup
In-Reply-To: <1238318710.23852.38.camel@twins>
References: <20090328194359.426029037@chello.nl>
	<20090328194929.451591360@chello.nl>
	<18894.48499.125187.92480@cargo.ozlabs.ibm.com>
	<1238318201.23852.17.camel@twins>
	<1238318710.23852.38.camel@twins>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 828
Lines: 19

Peter Zijlstra writes:

> In a perfect world, I'd introduce a self-ipi on UP and use that. Also,
> in that same perfect world, all arches would support cmpxchg()/xchg() so
> we could put the whole thing in generic code.

We already require atomic64_t, which not all architectures have - in
particular, ppc32 doesn't have it, though it does have cmpxchg/xchg
for 32-bit quantities.  The list of architectures that have atomic64_t
seems to be a proper subset of the list of architectures that have
cmpxchg, from a quick grep.  So I would think you can use cmpxchg if
necessary.

Paul.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/